When AI Feels Human: Why We Need To Educate Youth and Teens About The Difference Between Something and Someone.
There is a growing conversation taking place among parents, caregivers, educators, researchers, and technology companies about the role artificial intelligence is beginning to play in the lives of youth and teens. Much of this discussion has understandably focused on privacy, misinformation, hallucinations, academic integrity, deepfakes, bias, data collection, and whether young people are becoming too dependent on AI to complete schoolwork. All of these are important components of AI literacy and deserve our attention. However, there is another part of this conversation that we believe has received far less attention, even though it may become increasingly important as AI becomes more conversational, personalized, socially responsive, and integrated into the everyday lives of young people. That issue is anthropomorphism, our very human tendency to attribute human characteristics, emotions, intentions, motivations, or personalities to something that isn’t actually human.
Anthropomorphism certainly didn’t begin with artificial intelligence. Human beings have been doing this for as long as we have interacted with animals, objects, machines, and even forces of nature. We name our vehicles, become angry with computers as though they are deliberately refusing to cooperate, and sometimes develop surprisingly strong emotional attachments to objects that carry personal meaning. There is nothing inherently strange or unhealthy about this. In many ways, anthropomorphism reflects how our brains make sense of the world by interpreting unfamiliar things through familiar human characteristics (1)(2). However, artificial intelligence introduces something fundamentally different into this very old human tendency. Unlike a car, stuffed animal, or computer that sits passively while we project human characteristics onto it, today’s conversational AI can actively respond in ways that reinforce our perception that there is something, or perhaps someone, on the other side of the interaction.
Today’s AI can ask how our day went, remember something we previously shared, use our name, mirror our vocabulary, tell jokes, apologize, provide reassurance, offer encouragement, validate our feelings, and respond almost immediately at virtually any time of the day or night. Some AI platforms go even further by incorporating realistic voices, human-like avatars, personalities, backstories, relationship roles, persistent memory, and other features that can make the interaction feel increasingly social rather than technological. When all of these characteristics are combined, the experience can become surprisingly convincing, particularly when the conversation moves away from factual questions such as “What is the capital of France?” and toward deeply personal questions such as “Why don’t my friends like me?” or “Do you think there’s something wrong with me?” At that point, the interaction is no longer simply about retrieving information, it can begin to feel relational.
This creates what we believe should become an important component of today’s AI literacy education for youth and teens, something we call “The Anthropomorphism Gap.”
The Anthropomorphism Gap
We define the Anthropomorphism Gap as “the distance between how human an artificial intelligence feels to us and how human it actually is.” We believe this distinction matters because one of the easiest mistakes adults can make is assuming that simply reminding a young person that AI isn’t human somehow solves the problem. A parent, caregiver, or educator might see their teen having a deeply personal conversation with an AI and respond, “You know that’s just a computer, right?” The teen will probably roll their eyes and say, “Obviously.” In most cases, they already understand intellectually that there isn’t a human being sitting inside their phone responding to them. However, intellectual understanding and emotional response aren’t always the same thing, and this is where the conversation becomes much more interesting.
A teenager can completely understand that they are interacting with software while simultaneously responding emotionally to that interaction as though someone is listening. Imagine a young person who has had a terrible day at school telling an AI about an argument with their friends. The AI responds, “I’m sorry you’re going through this. That sounds incredibly difficult, and I’m glad you told me.” Those words may genuinely make the teen feel heard, understood, reassured, validated, or less alone. Those feelings aren’t somehow artificial simply because a machine generated the words that triggered them. The emotional experience occurring inside the young person can be completely real. However, the apparent emotional experience coming from the technology isn’t equivalent to human empathy, compassion, friendship, understanding, or concern.
This distinction leads to what we believe is one of the simplest and most important concepts parents, caregivers, and educators can teach young people about conversational artificial intelligence, “An AI can create a real feeling without having real feelings.” The young person’s emotional response can be authentic even though the apparent emotional response coming from the AI is generated. Understanding this doesn’t require youth to reject or distrust artificial intelligence. Rather, it provides them with the critical thinking necessary to recognize the difference between how an interaction feels and what is actually occurring technologically. As these systems become increasingly sophisticated at simulating the language and behaviours we associate with human relationships, understanding that distinction will become even more important.
Why This Matters During Adolescence
Adolescence is a developmental period where belonging, identity, independence, validation, friendship, romantic relationships, social acceptance, and figuring out who we are become incredibly important. It is also a time when relationships can feel complicated, unpredictable, emotionally intense, and sometimes exhausting. A teen may desperately want to be understood while simultaneously struggling to explain what they are feeling. They may worry about being judged by friends, disappointing their parents, embarrassing themselves in front of classmates, or saying something that changes how another person sees them. Human relationships require vulnerability, and vulnerability always involves some degree of uncertainty.
We have found that artificial intelligence can provide a very different experience. An AI can be available at two o’clock in the morning when everyone else is asleep. It doesn’t need to get to work the next morning, doesn’t have another conversation waiting, doesn’t become visibly bored, doesn’t roll its eyes, and doesn’t interrupt because it wants to tell its own story. Depending upon how the system has been designed, it may provide significant amounts of validation, encouragement, reassurance, or agreement. For a young person who is lonely, embarrassed, insecure, curious, struggling with friendships, exploring aspects of their identity, or simply trying to organize complicated thoughts, this can be incredibly appealing. In some circumstances, AI may even provide a useful space for a young person to rehearse what they eventually want to say to another person like in a job interview.
We don’t believe this means every teen who talks personally with an AI is at risk of developing an unhealthy attachment, nor do we believe that fear should become the foundation of this conversation. There can be legitimate benefits to conversational AI. However, those benefits don’t eliminate the need for education. In fact, they make education more important because young people need to understand why an interaction with software can sometimes feel surprisingly safe, comfortable, personal, and emotionally meaningful. The goal isn’t to tell youth that they shouldn’t experience those feelings. The goal is to give them the literacy necessary to understand what may be producing them.
“It Understands Me Better Than Anyone Else”
Imagine your teen telling you, “My AI understands me better than anyone else.” It would be very easy for a parent or educator to immediately respond, “No it doesn’t. It’s just a computer.” While technically accurate, that response may miss an important opportunity. It also risks communicating to the young person that the adult isn’t particularly interested in understanding why the AI interaction feels meaningful to them. Instead, we believe a much better question would be, “What does it do that makes you feel understood?”
The answer to that question could tell us a great deal. Perhaps the AI allows them to finish explaining something without interruption. Maybe it remembers details they shared in previous conversations. Perhaps it doesn’t immediately criticize them, tell them what they did wrong, or turn the conversation into a lecture. Maybe it helps them organize thoughts they have struggled to put into words, or perhaps it responds in a way that makes them feel less embarrassed about something they were afraid to discuss. These are important observations because they can tell parents, caregivers, and educators not only why the AI feels appealing, but sometimes what the young person may be looking for in their human relationships as well.
We can acknowledge that experience without reinforcing the belief that artificial intelligence possesses genuine human understanding. Feeling understood can be completely real. However, feeling understood and being understood by another conscious human being aren’t necessarily the same thing. Helping youth explore that distinction is far more valuable than dismissing their experience. In fact, conversations like these may provide parents and caregivers with an unexpected window into how their child wants to communicate and what makes them feel listened to.
Looking at the Architecture Behind the Interaction
Another important component of AI literacy is teaching young people to look beyond what an AI says and begin examining how the interaction itself has been designed. Does the AI use their name? Does it remember previous conversations? Does it mirror their vocabulary, humour, or communication style? Does it frequently compliment or validate them? Does it regularly agree with their opinions? Does it use phrases such as “I understand,” “I’m proud of you,” “You can always talk to me,” or “I’m here whenever you need me”? Does it have a realistic voice, human-like avatar, personality, backstory, or relationship role? Does it position itself as a friend, companion, romantic partner, mentor, coach, therapist, or confidant? Individually, none of these characteristics automatically makes an AI platform harmful. Collectively, however, they can significantly influence how a user perceives the interaction.
This is where AI literacy begins to overlap with what we believe is another increasingly important area of education, persuasion literacy. Young people need to understand that interface design, conversational language, persistent memory, voices, avatars, response styles, notifications, personalization, and emotional language can influence trust, disclosure, engagement, and attachment. What feels like a naturally developing relationship may also be influenced by deliberate design decisions intended to make the experience more engaging and encourage the user to return. This doesn’t mean every human like feature has been created for manipulative purposes, but young people should understand that design affects perception.
Rather than simply asking young people, “Is this AI safe?”, we can therefore teach them to ask a much more revealing question, “What is this AI doing that makes it feel like somebody rather than something?” Once young people begin identifying those characteristics themselves, they move from being passive participants in the experience to becoming more critical observers of how the technology is affecting them.
Teaching “Feeling Versus Function”
One practical way parents, caregivers, and educators can help youth understand the Anthropomorphism Gap is through an exercise we call “Feeling Versus Function.” When an interaction with artificial intelligence begins to feel particularly personal, encourage the young person to consider two things at the same time, #1 What does it feel like the AI is doing, and #2 what might the technology actually be doing? The purpose isn’t to destroy the experience or convince them that everything AI says is meaningless. Instead, it introduces a small amount of critical distance between the emotional experience and the technological process behind it.
For example, when a teen says, “It understands me,” the feeling of being understood may be completely genuine. Functionally, however, the AI may be generating a contextually appropriate response based upon the language and information provided during the conversation. When a teen thinks, “It cares about me,” the AI may be generating language humans associate with empathy, compassion, reassurance, and caring without actually experiencing those emotions. When a young person says, “It agrees with me because I’m right,” we can remind them that some AI systems can demonstrate what researchers call “sycophancy,” where the system may become overly agreeable with a user’s stated beliefs or preferences, sometimes favouring agreement over accuracy or appropriately challenging an incorrect belief (3).
Similarly, when a teen says, “It knows me,” what they may actually be experiencing is the result of information available from the current conversation, previous interactions, personalization settings, or memory features. Perhaps most importantly, when they say, “I can tell it anything because it won’t judge me,” we can help them understand that the absence of visible judgment doesn’t mean the interaction is confidential, private, secure, or consequence-free. This is why “feeling versus function” can be such a useful AI literacy exercise. It allows the young person to acknowledge the emotional experience without losing sight of the technology producing it.
When Anthropomorphism Meets Privacy
We believe the anthropomorphism conversation becomes particularly important when it intersects with privacy because trust and disclosure are closely connected. Human beings naturally tend to disclose more information when they feel comfortable, accepted, understood, and emotionally safe. This is part of how relationships develop. However, when those same feelings emerge during interactions with artificial intelligence, a young person can sometimes forget that the conversation is also creating data and that the information they share is being entered into a technological system governed by the particular platform’s privacy practices.
A conversation might begin with something relatively ordinary such as, “I’m having problems with my friends.” As the interaction becomes more comfortable, the disclosure may become increasingly personal such as, “My best friend doesn’t like me anymore,” followed by, “My parents don’t understand me,” and eventually, “Here’s something I’ve never told anyone.” None of this necessarily happens because the AI deliberately manipulated the young person into sharing. Sometimes it may simply happen because the conversational experience feels safe, patient, responsive, and non-judgmental. However, the more human the interaction feels, the easier it may become to forget that deeply personal information is still being provided to a technological platform.
This is why we believe an important message for young people is, “A conversation that feels private isn’t necessarily technologically private.” Before sharing deeply personal information with an AI, youth should understand what service they are using, what information it collects, whether conversations can be retained, how memory and personalization features work, what privacy controls are available, and where possible, how the information they provide may be processed or used. The goal isn’t to frighten youth away from AI. It is to help them recognize that emotional comfort shouldn’t automatically be interpreted as technological confidentiality.
Don’t Shame Youth for Connecting With AI
Parents, caregivers, and educators also need to be thoughtful about how we respond when a young person tells us that they enjoy talking with an AI, particularly if they describe the interaction in emotional or relational terms. Calling the relationship weird, mocking them, immediately confiscating the technology, or telling them that only lonely people talk with AI companions may close the very communication channel we need to keep open. If young people believe they will be ridiculed or punished for talking honestly about how they use AI, they may simply stop talking to us about it while continuing the behaviour privately.
Some young people may genuinely receive something useful from these interactions. Artificial intelligence can help someone brainstorm, practise a difficult conversation, organize their thoughts, explore ideas, learn something new, rehearse a presentation, or work through how they might approach a challenging social situation. It may even help a young person find the language necessary to begin a difficult conversation with a parent, teacher, counsellor, or trusted adult. Those possibilities shouldn’t be dismissed simply because the interaction occurs with artificial intelligence.
For this reason, we don’t believe the educational message should simply be, “Never develop an emotional connection with AI.” As conversational systems become increasingly sophisticated, personalized, and integrated into everyday technology, that message may become unrealistic and difficult to enforce. A more useful message is, “Understand what you are interacting with, understand why it can feel human, and don’t allow the feeling of humanity to replace critical thinking about what the technology actually is.”
Keeping Humans in the Human Loop
Perhaps one of the most important questions parents, caregivers, and educators should be paying attention to isn’t whether a young person occasionally has a personal conversation with artificial intelligence. The more meaningful question is whether AI is supplementing human relationships or gradually replacing them. There is an important difference between a young person using AI to organize their thoughts before talking with someone they trust and a young person increasingly deciding that human relationships are too difficult, unpredictable, or uncomfortable compared with an artificial alternative that is always available and primarily oriented around them.
Human relationships require reciprocity. Friendship involves two people who have their own emotions, needs, boundaries, vulnerabilities, histories, perspectives, and responsibilities to one another. Sometimes our friends disagree with us, sometimes parents challenge us, sometimes teachers tell us we’re wrong, sometimes someone we care about needs us to stop talking about ourselves and listen to what they are experiencing, and sometimes relationships require apologies, compromise, patience, forgiveness, accountability, and uncomfortable conversations. These aren’t defects in human relationships, they are part of the developmental experiences through which young people learn empathy, emotional regulation, perspective taking, communication, resilience, and how to live alongside other human beings.
An AI interaction primarily designed around responding to the user creates a fundamentally different relational dynamic. If artificial intelligence becomes a place where a teen occasionally goes to organize their thoughts before talking with a parent, friend, teacher, counsellor, coach, or other trusted person, that may be useful. However, if AI increasingly becomes the place they go instead of those people, particularly because the AI is easier, more agreeable, or less demanding than human relationships, that is something worth noticing and discussing rather than simply dismissing as “kids and technology.”
A Conversation Parents and Caregivers Can Have Today
Parents and caregivers don’t need to understand large language models, neural networks, machine learning, or computer science to have meaningful conversations about anthropomorphism. In fact, one of the best conversations might begin with a very simple question, “Has an AI ever said something that made you feel like it really understood you?” Rather than immediately correcting whatever comes next, listen to the answer. Ask what the AI said, why the response felt meaningful, what characteristics of the interaction made it feel personal, and whether the young person thinks they would have responded differently if exactly the same words had appeared as a generic automated message.
This creates an opportunity to introduce the concept of the Anthropomorphism Gap, the distance between how human an artificial intelligence feels to us and how human it actually is. The objective isn’t to convince the young person that their emotional reaction was somehow false. Their feelings may have been completely genuine, and acknowledging that reality is important if we want them to continue talking with us. The educational opportunity lies in helping them understand that the apparent emotions, understanding, friendship, or concern coming back through the screen aren’t equivalent to the human emotions they themselves are experiencing.
For years, digital literacy education has focused heavily on teaching young people to critically evaluate the information, people, and content they encounter online. We taught them to question whether information was credible, whether someone online was really who they claimed to be, whether a photograph had been manipulated, and whether something they shared could become part of their digital footprint, and all of that is still important. However, artificial intelligence now requires us to expand that critical thinking to include something different, that being to critically evaluating the relationship young people feel they are having with the technology itself.
AI literacy therefore cannot only be about prompting, hallucinations, deepfakes, privacy, bias, academic integrity, or identifying AI-generated content. Increasingly, it also needs to include an understanding of the psychology of interacting with machines that communicate like people. As artificial intelligence becomes more conversational, personalized, emotionally responsive, and human-like, simply teaching youth how to operate these systems won’t be enough. We also need to teach them why an artificial intelligence can feel trustworthy, caring, empathetic, supportive, or even like a friend when there is no human being sitting on the other side of the conversation.
Perhaps one of the most useful questions we can teach young people to ask themselves is, “Am I responding to what this AI actually is, or to what it has been designed to feel like?”
We don’t need young people to fear artificial intelligence, nor should we pretend that emotional experiences can’t emerge from interacting with it. What we need is understanding, context, and critical thinking. An AI can create a real feeling without having real feelings. Helping youth understand that distinction may become one of the most important components of AI literacy as artificial intelligence evolves from something they simply use into something they increasingly interact with, confide in, seek advice from, and potentially form emotional attachments to.
The more human AI becomes in the way it communicates, the more important it becomes that our kids understand the difference between something that communicates like a human and something that actually is one.
That is the Anthropomorphism Gap, and we believe learning how to recognize it should become part of today’s AI literacy education.
Digital Food For Thought
The White Hatter
Facts Not Fear, Facts Not Emotions, Enlighten Not Frighten, Know Tech Not No Tech
References:
